Most image compression algorithms tend to decorrelate the RGB color components as part of the compression process. Nevertheless, the high correlation between color components could be helpful if mutual spectral information is exploited in the compression process. In this paper we introduce a new approach to color image compression that increases and exploits the high mutual spectral information of RGB components to improve the results' quality. Simulation results are presented and discussed. Our conclusion is that higher correlation between color components could contribute to improved image compression and be superior to the common approach of decorrelated image components.